Parking space remote monitoring management system based on RFID technology

Abstract

The invention relates to the technical field of parking space monitoring, and discloses a parking space remote monitoring management system based on an RFID technology, and the system comprises a sensing layer which is composed of ultrahigh frequency RFID reader-writers, ultrasonic sensors and geomagnetic sensors which are distributed at all parking spaces of a parking lot; the ultrahigh frequency RFID reader-writer is used for identifying an RFID tag carried by a vehicle to obtain vehicle identification information; the ultrasonic sensor detects the distance of an obstacle above a parking space through a sound wave reflection principle so as to judge the vertical occupation condition of the parking space. The geomagnetic sensor is used for detecting the magnetic field intensity change of a parking space area according to the disturbance of an iron-containing part of a vehicle to a geomagnetic field so as to judge the horizontal occupation condition of the parking space; and the transmission layer adopts a wired and wireless combined mode. By adopting the technical scheme of the sensing layer integrating the ultrahigh-frequency RFID reader-writer, the ultrasonic sensor and the geomagnetic sensor, the technical effect of collecting the multi-dimensional information of the parking space is achieved, and the defects that the parking space state judgment is inaccurate and is easily interfered by the environment are overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of parking space monitoring, and in particular to a parking space remote monitoring and management system based on RFID technology. Background Art

[0002] With the acceleration of urbanization, the number of cars has continued to rise, and the problem of parking difficulties in cities has become increasingly prominent. Efficient and accurate remote monitoring and management of parking spaces has become an important issue that needs to be solved urgently. In this context, the existing parking space monitoring and management technology has gradually exposed many shortcomings, as follows: When collecting parking space information, most existing technologies use a relatively single sensor solution or a simple sensor combination. For example, some parking lots rely solely on ultrasonic sensors to determine the status of parking spaces. The principle is to measure the distance through sound wave reflection to infer whether the parking space is occupied. However, this single sensor solution has obvious limitations. On the one hand, when encountering irregularities on the surface of the vehicle, the path and time of sound wave reflection will be affected, resulting in deviations in the measured distance and failing to accurately reflect the actual occupancy of the parking space; on the other hand, large changes in weather temperature will also interfere with ultrasonic sensors, because temperature will change the speed of sound wave propagation and the performance parameters of the sensor itself, thereby increasing the measurement error, making the judgment of the parking space status inaccurate and susceptible to interference from environmental factors.

[0003] Traditional parking systems often use a single wired or wireless communication method for data transmission. For example, some parking lots only use Wi-Fi communication to transmit data. During peak hours of vehicle entry and exit, a large amount of vehicle information needs to be transmitted in real time, such as vehicle entry and exit records, parking space status updates, etc. At this time, data traffic increases sharply, and Wi-Fi networks are prone to insufficient bandwidth, which leads to packet loss, delays and other problems, seriously affecting the reliability of data transmission. Once data congestion or interruption occurs, the parking lot management end cannot obtain accurate parking space information in a timely manner, and cannot effectively guide and manage vehicles. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a parking space remote monitoring and management system based on RFID technology, which solves the problem that only relying on a single sensor cannot obtain multi-dimensional information of the parking space and cannot fully and accurately grasp the actual usage status of the parking space.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A parking space remote monitoring and management system based on RFID technology, comprising: The sensing layer consists of UHF RFID readers, ultrasonic sensors, and geomagnetic sensors distributed in each parking space in the parking lot; The UHF RFID reader is used to identify the RFID tag carried by the vehicle to obtain the vehicle identification information; The ultrasonic sensor detects the distance of obstacles above the parking space through the principle of sound wave reflection to determine the vertical occupancy of the parking space; The geomagnetic sensor is used to detect the change of magnetic field intensity in the parking area according to the disturbance of the iron-containing parts of the vehicle to the geomagnetic field to determine the horizontal occupancy of the parking space; The transport layer uses a combination of wired and wireless methods, including industrial Ethernet, Wi-Fi, and 4G / 5G networks, to transmit the data collected by the perception layer to the application layer; The application layer includes the parking lot management server, database, mobile phone APP for car owners, and display terminals in the parking lot; The parking lot management server is used to receive data from the transmission layer and run data fusion, analysis and decision-making algorithms to perform intelligent management of parking spaces; The database is used to store vehicle information, parking space status historical data, and user information, and the mobile phone APP is used to provide parking space query, reservation, and navigation services for car owners; The display terminal is used to display the overall situation of parking spaces in the parking lot in real time to guide car owners to park.

[0006] Preferably, the UHF RFID system has a built-in large-capacity storage chip that can record the vehicle's license plate number, vehicle model, owner's contact information and entry time information, and the chip has an adaptive tuning function.

[0007] Preferably, the temperature compensation circuit built into the ultrasonic sensor automatically adjusts the measurement parameters according to the ambient temperature.

[0008] Preferably, the geomagnetic sensor is a three-axis geomagnetic sensor with a measurement sensitivity of 0.1nT. A high-precision ADC is integrated inside the sensor to convert the magnetic field strength into a digital signal output, and the sensor is equipped with a multi-layer magnetic shielding cover outside.

[0009] Preferably, in the sensing layer, the RFID reader is installed on one side of the parking space at a height of 2-2.5 meters from the ground, and the antenna is tilted 30°-45° toward the parking space; the ultrasonic sensor is installed 0.8-1.2 meters directly above the parking space; the geomagnetic sensor is buried in the center of the parking space ground at a depth of 5-10 centimeters, and is covered with a layer of electromagnetic shielding and heat insulation composite layer.

[0010] Preferably, the parking lot management server adopts a multi-data deep fusion algorithm based on an improved Bayesian network and DS evidence theory, including a data acquisition and preprocessing module, a Bayesian network reasoning module, and a DS evidence theory fusion module; The data acquisition and preprocessing module uses an STM32F4 microcontroller to build a circuit, synchronously triggers each sensor to collect data through a timer, and preprocesses the collected data using a median filter, a Kalman filter, and a CRC check algorithm; The Bayesian network reasoning module constructs a Bayesian network containing four layers of nodes, the first layer is the sensor observation node, the second layer is the environmental impact node, the third layer is the parking space status intermediate node, and the top layer is the parking space real status node; The DS evidence theory fusion module performs DS evidence theory fusion after the Bayesian network inference obtains the preliminary judgment of each sensor on the parking space status.

[0011] Preferably, the transmission layer adopts multi-link redundant communication, deploys industrial Ethernet as the backbone network in the parking lot, connects various data collection nodes with the management server, and deploys Wi-Fi6 wireless network to cover the entire parking lot.

[0012] Preferably, in the application layer, the parking lot management server adopts RabbitMQ message queue technology to push information including parking space status updates and vehicle entry and exit records to relevant terminals.

[0013] Preferably, the system also includes an intelligent self-repair and adaptive adjustment module, which starts an intelligent diagnosis program based on machine learning every 3 minutes, collects data from each sensor for the past 10 minutes, including features such as measurement values, data fluctuation range, and signal strength, and inputs the data into a pre-trained decision tree model to determine whether the sensor is working properly.

[0014] Preferably, the system also includes a remote collaborative operation and maintenance module, which uses a microservice architecture to perform cluster analysis and correlation analysis on the data of multiple parking lots of the same type.

[0015] The present invention provides a parking space remote monitoring and management system based on RFID technology. It has the following beneficial effects: 1. The present invention adopts a perception layer technology solution that integrates ultra-high frequency RFID readers, ultrasonic sensors and geomagnetic sensors to achieve the technical effect of accurately collecting multi-dimensional information of parking spaces. Compared with the technical solutions in the prior art that only rely on a single type of sensor or a simple sensor combination, the present invention solves the shortcomings of inaccurate parking space status judgment and susceptibility to environmental interference. The present invention uses the coordination of three sensors, utilizes their respective advantages, and combines intelligent algorithms to process data, so as to stably and accurately judge the parking space occupancy status under complex working conditions.

[0016] 2. The present invention adopts a transmission layer technical solution that combines wired and wireless and has multi-link redundant communication, which achieves the technical effect of ensuring stable and high-speed data transmission. Compared with the technical solution of a single wired or wireless communication method in the prior art, it solves the shortcomings of poor transmission reliability and easy data congestion or interruption.

[0017] 3. The present invention adopts an application-layer technical solution of a multi-data deep fusion algorithm based on an improved Bayesian network and DS evidence theory, which achieves the technical effect of improving the accuracy of parking space status judgment. Compared with the technical solution based on simple threshold judgment or a single algorithm in the prior art, it solves the shortcomings of low judgment accuracy and inability to integrate multi-source information for effective decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a diagram of the architecture of the monitoring and management system of the present invention; Figure 2 This is an architecture diagram of the perception layer in the present invention; Figure 3 This is an architecture diagram of the transport layer in the present invention; Figure 4 This is an architecture diagram of the application layer in the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Please see attached Figure 1 - Attachment Figure 4 The embodiment of the present invention provides a parking space remote monitoring and management system based on RFID technology, including: The sensing layer consists of UHF RFID readers, ultrasonic sensors, and geomagnetic sensors distributed in each parking space in the parking lot; The UHF RFID reader is used to identify the RFID tag carried by the vehicle to obtain the vehicle identification information; The ultrasonic sensor detects the distance of obstacles above the parking space through the principle of sound wave reflection to determine the vertical occupancy of the parking space; The geomagnetic sensor is used to detect the change of magnetic field intensity in the parking area based on the disturbance of the vehicle's iron-containing parts to the geomagnetic field to determine the horizontal occupancy of the parking space; Specifically, the directional antenna of the UHF RFID reader has a high gain characteristic of 15dBi, which can highly focus the radio frequency signal and strongly project it to the parking area, aiming to enhance the recognition efficiency of the tag. In one embodiment, when a vehicle carrying an RFID tag enters the predetermined signal coverage area around the parking space, the reader immediately starts the recognition process, establishes a close connection with the high-speed data acquisition circuit built on the STM32F4 microcontroller through the SPI interface, and quickly captures the tag information and monitors the signal strength indication in real time at a rate of up to 50Mbps. In the standard test scenario simulated in the laboratory, when the reader is 3 meters away from the tag and is in an ideal environment without external interference, the reader can accurately identify more than 500 tags per second. At the same time, the measurement error of RSSI is strictly controlled within a very small range of ±2dBm. This is due to the built-in precision calibration module of the system, which periodically compares with the pre-set standard signal source and fine-tunes the measurement parameters in real time to ensure that the RSSI data obtained is true and reliable, laying a solid foundation for the subsequent accurate judgment of the parking space status. In addition, in the face of complex and changeable electromagnetic environments, such as sudden stray signals from surrounding communication base stations or electromagnetic leakage from electrical equipment in parking lots, the spectrum monitoring module built into the reader can sense the frequency band interference at a millisecond response speed. Once it is determined that the interference intensity of the current frequency band exceeds the preset threshold, this threshold is set based on a large number of field tests and electromagnetic theory analysis, covering multi-dimensional considerations such as the proportion of noise components and signal attenuation amplitude. The reader immediately drives the frequency hopping command and seamlessly switches to a relatively clean frequency band for continuous operation, ensuring the stability and continuity of the tag reading operation. Under the periodic excitation of the driving circuit, the high-performance piezoelectric ceramic transducer built into the ultrasonic sensor transmits ultrasonic pulses to the space above the parking space at preset time intervals. In one embodiment, these pulses are reflected when encountering obstacles such as vehicles. The receiving circuit of the sensor quickly captures the reflected echo, and then efficiently converts the sound wave reflection time into a distance value through the ADC conversion module, so as to accurately infer the vertical occupancy details of the parking space. For example, multiple repetitive measurement experiments were carried out on simulated target objects at a standard distance of 1.5 meters, and the measurement accuracy was always stably maintained within the strict range of ±1.5cm. Its built-in intelligent temperature compensation circuit, which cleverly integrates thermistors, can sense the ambient temperature in real time. Relying on the temperature-accuracy calibration table pre-stored in the chip and deeply fitted by massive experimental data, it dynamically and adaptively adjusts the sound wave emission and reception parameters to ensure that under different temperature conditions, covering common ranges such as -10℃, 25℃, and 50℃, the fluctuation amplitude of the measured distance is strictly controlled within the threshold of ±0.5cm, providing vertical occupancy information support for parking space status judgment; The core component of the geomagnetic sensor is surrounded by a high-permeability multi-layer magnetic shield, which is made of Permalloy and can effectively reduce the external stray magnetic field by more than 90%, making the low-frequency magnetic field generated by common interference sources such as nearby power cables and transformers almost impermeable. As an option, its internal integrated 12-bit ADC converts the changes in the geomagnetic field intensity into digital signal output, which facilitates the subsequent data processing process. In one embodiment, the sensor transmits three-axis magnetic field intensity data (in nT, with an accuracy of 0.1nT) to the STM32F4 microcontroller through the I2C interface every 0.1 second. The microcontroller properly stores this data according to the preset 16-bit quantization accuracy standard, and combines the internal clock to mark the data acquisition timestamp to ensure the synchronization and integrity of the data in all aspects. For example, in the test phase, simulated vehicle models with different iron contents were placed to test their performance. Even in the case of models with very low iron content, the sensor was still able to detect changes in magnetic field strength with extremely high accuracy, and the measurement error was controlled within an ultra-narrow range of ±0.1nT. Moreover, in the case of intentionally set external magnetic field interference scenarios, such as placing strong magnets to simulate high-intensity interference sources, the sensor was still able to accurately measure and output correct magnetic field strength data, and the error was always within the allowable range, providing a reliable basis for accurately judging the horizontal occupancy of parking spaces, and becoming an indispensable key link in the parking space status determination system. Through the sensors in the perception layer, the system acquires basic data related to parking spaces from all directions and angles, laying a solid foundation for key links such as subsequent data transmission, deep fusion processing, and accurate determination of parking space status. The transport layer uses a combination of wired and wireless methods, including industrial Ethernet, Wi-Fi, and 4G / 5G networks, to transmit the data collected by the perception layer to the application layer; Specifically, the industrial Ethernet is used as the backbone network of the transmission layer, which is built according to the IEEE802.3 standard and uses Category 6 or Category 6A twisted pair cables as the transmission medium. Specifically, each data acquisition node, that is, the node where each sensor of the perception layer is located, is connected to the Ethernet switch through a standard RJ45 interface. The switch plays a core role in aggregating and forwarding data. In one embodiment, when many vehicles enter and exit the parking lot at the same time, and each sensor collects and transmits data at a high frequency, the industrial Ethernet can transmit a large amount of data to the management server with its stable physical link and efficient switching mechanism, ensuring that each data packet can reach the destination in a very short time according to the established routing rules, avoiding data loss or delay caused by network congestion, and providing comprehensive and accurate data support for subsequent parking space management decisions in a timely manner; In one embodiment, the Wi-Fi 6 wireless network is responsible for covering the entire parking lot, providing flexible data transmission channels for mobile devices and some temporarily accessed nodes. It uses advanced OFDMA technology, and can intelligently allocate channels in the 2.4GHz and 5GHz frequency bands according to the actual channel resource occupancy, so as to achieve simultaneous parallel communication of multiple devices. Specifically, multiple wireless access points (APs) use Mesh networking to automatically establish reliable wireless links with each other, so that a seamless roaming network environment is formed in the entire parking lot, eliminating signal dead spots. In one embodiment, when parking lot staff use handheld inspection terminals to inspect different areas, or when temporary debugging equipment needs to access the network, the Wi-Fi 6 network can respond quickly and allocate independent channel resources to these mobile terminals to ensure their stable connection and smooth data transmission. Even when the equipment is densely populated in the local area, it can maintain good network performance by dynamically adjusting bandwidth, optimizing modulation and demodulation methods, etc., to meet diverse access needs, further enrich the data source channels of the system, and enhance the overall flexibility and adaptability; In some areas with dense vehicles and peak data traffic such as parking lot entrances and exits, the present invention adds 5G network links as a supplement. With its outstanding advantages of low latency and high bandwidth, the 5G network can easily cope with the sudden large data transmission needs in these areas. Specifically, through deep integration with the operator's network, based on the 5G network slicing technology, a dedicated network slice is opened for the parking space management system to ensure that the data transmission of this system enjoys high priority and is not interfered by other non-critical business traffic. In one possible implementation, when vehicles frequently enter and exit the parking lot entrance and exit, and large-capacity data including high-definition license plate recognition images and accurate parking space occupancy information need to be transmitted in real time, the 5G network link can quickly undertake and efficiently transmit these data to ensure the real-time and integrity of the data. In addition, an intelligent routing algorithm is deployed on the server side, which will comprehensively monitor the status of each link at extremely short time intervals. The monitoring indicators involved include key parameters such as bandwidth utilization, delay jitter, and packet loss rate. Once a link is found to be faulty, such as a switch port in the industrial Ethernet is damaged, causing the packet loss rate of the link to instantly soar to 100%, or performance degradation, such as too many access devices in the area where a Wi-Fi6 AP is located, bandwidth utilization exceeding 90%, and latency exceeding 50ms, the intelligent routing algorithm will immediately switch data transmission to the optimal backup link based on the preset decision logic, taking into account multiple factors such as the historical performance, current load, and business needs of each link. The entire switching process is fast and smooth, which maximizes the continuity of data transmission and ensures that parking space-related data can be transmitted to the application layer without hindrance for subsequent processing; Through the design and efficient operation of the transport layer in step S2, different links cooperate and complement each other, building a reliable and high-speed data transmission network for the system, ensuring that the parking space data collected by the perception layer can reach the application layer accurately, laying a solid foundation for realizing the intelligent management function of the entire parking space remote monitoring and management system.

[0021] The application layer includes the parking lot management server, database, mobile phone APP for car owners, and display terminals in the parking lot; The parking lot management server is used to receive data from the transmission layer and run data fusion, analysis and decision-making algorithms to manage parking spaces intelligently; The database is used to store vehicle information, parking space status historical data, and user information, and the mobile APP is used to provide parking space query, reservation, and navigation services for car owners; The display terminal is used to display the overall situation of parking spaces in the parking lot in real time to guide car owners to park; Specifically, the parking space status determination based on the multivariate data fusion algorithm: In general, the parking lot management server, as the core device of the application layer, is equipped with a high-performance multi-core processor and a large-capacity memory (not less than 32GB), and runs a parking space management software customized and developed based on the Linux operating system. Its core algorithm is a multivariate data deep fusion algorithm based on the improved Bayesian network and DS evidence theory. Specifically, first in the data receiving link, the server will parse and verify the sensor data from the transmission layer in accordance with strict data formats and communication protocols to ensure the integrity and accuracy of the data. In a possible implementation method, the tag information received from the ultra-high frequency RFID reader, the distance measurement value of the ultrasonic sensor, and the magnetic field strength data of the geomagnetic sensor will be stored in the corresponding cache area respectively, and preliminary data cleaning will be performed to eliminate obviously abnormal data points, such as data beyond the normal value range, such as negative or extremely large values ​​of ultrasonic measurement distance, which do not conform to physical logic.

[0022] Next, enter the data fusion stage. The constructed Bayesian network consists of four layers of nodes. The first layer is the sensor observation nodes, namely the RFID tag recognition result (S1), the ultrasonic distance measurement (S2), and the geomagnetic measurement of the magnetic field intensity change (S3). The second layer is the environmental impact nodes, including the electromagnetic interference intensity (E1), temperature (E2), and humidity (E3). The third layer is the intermediate node of the parking space status, divided into empty parking space (O1), partially occupied (O2), and fully occupied (O3). The top layer is the node of the true parking space status (T), with values of occupied (Y) or unoccupied (N). Directed connections are established between the nodes based on physical principles and practical experience. For example, the electromagnetic interference intensity (E1) directly affects the RFID tag recognition result (S1) because strong electromagnetic interference will reduce the reliability of communication between the tag and the reader, increasing the probability of recognition errors. Temperature (E2) not only affects the geomagnetic measurement accuracy (by changing the characteristics of the magnetic materials inside the sensor) but also indirectly affects the node of the parking space occupancy level. Considering that high temperatures may cause thermal expansion and contraction of vehicle components, affecting the actual occupancy of the vehicle in the parking space. Using a large amount of historical data, covering the operation data of the parking lot in different seasons, weather, and time periods, the maximum likelihood estimation method is used to learn the conditional probability distribution between the nodes. Taking the conditional probability (PS1 = 1|E1 = i) as an example, that is, the probability of successful RFID tag recognition when the electromagnetic interference intensity is (i). By counting the proportion of the actual number of successful RFID tag recognitions to the total number of measurements at different electromagnetic interference intensity levels (classified according to the pre-set interference intensity thresholds, such as divided into low, medium, and high levels, corresponding to (i = 1, 2, 3) respectively), this is used as a reliable estimate of this conditional probability. Similarly, for complex conditional probabilities, such as (PS2 < d|O2) (the probability that the ultrasonic measurement distance is less than the threshold (d) when the parking space is partially occupied), (PS3 = m|T = Y, E2 = t) (the probability that the geomagnetic measurement magnetic field intensity change is (m) when the true parking space status is (Y) and the temperature is (t)), etc., they are all analyzed and estimated in depth in strict accordance with the corresponding data statistical rules, gradually constructing a complete and accurate conditional probability table, providing a solid basis for the efficient inference of the subsequent Bayesian network; After the Bayesian network initially infers and outputs the preliminary judgment results of each sensor on the status of the parking space, these results are cleverly transformed into independent evidence sources, and the basic probability assignment (BPA) is reasonably allocated based on multi-dimensional factors. Assume that the sensor set is (S={s_{1},s_{2},s_{3}}) (accurately corresponding to RFID, ultrasonic, and geomagnetic sensors, respectively). For the key proposition (A) that the parking space is occupied, the sensor's historical accuracy (Accuracy_{s_{i}}, obtained through long-term historical data statistics, such as the average accuracy of the past year), current working status (Status_{s_{i}}, based on the feedback information of the sensor's built-in self-test program, such as whether the ultrasonic sensor beam is normal, whether the RFID reader has internal error codes, etc.) and consistency with other sensor data (Consistency_{s_{i}}, measured by calculating the correlation coefficient or similarity with other sensor data, such as using the Pearson correlation coefficient method) and other key factors are carefully determined to determine its BPA. For example, for RFID sensors, the BPA calculation formula is designed as:

[0023] in, is the weight coefficient, the value range is strictly limited to between 0 and 1, and satisfies Through repeated experiments and data analysis, the weight coefficients are reasonably determined to ensure that the BPA allocation can reflect the historical performance of the sensor and take into account the current actual working conditions. and And each sensor for the empty parking space proposition The BPA of the data sets lays the foundation for subsequent fusion. Then, the DS synthesis rule is used to deeply fuse multiple evidence sources to obtain a more reliable and accurate comprehensive judgment on the parking space status. , The BPA of two sensors for proposition (A) are as follows:

[0024] First, synthesize the BPA of RFID and ultrasonic sensor to obtain new BPA , and then BPA with geomagnetic sensor (i.e. Perform secondary synthesis to obtain the final confidence level about the parking space being occupied Through multiple such fusion operations, the judgment information of each sensor can be fully integrated, the limitations of a single sensor can be overcome, and the accuracy of parking space status judgment can be improved. According to actual tests, in a variety of unfavorable conditions such as complex electromagnetic environments and irregular parking of vehicles, the accuracy of judging whether a parking space is occupied by relying solely on a single sensor may be around 70%-80%. After fusion through DS evidence theory, the accuracy of parking space status judgment can be increased to more than 90%, effectively enhancing the reliability of the system's judgment of parking space status.

[0025] In this embodiment, the application layer provides various information feedback channels for different users to meet the needs of all parties for parking space information. For parking lot managers, various statistical reports can be queried at any time through the server-side database. These reports cover key indicators such as parking space utilization rate, income, and failure rate, which is convenient for managers to fully understand the operation status of the parking lot, discover potential problems in time, and make decision adjustments.

[0026] For the majority of car owners, they mainly use mobile APP to obtain parking space-related information and enjoy convenient parking services. The mobile APP integrates map navigation and data interaction function modules. Specifically, it uses Firebase Cloud Messaging to push notifications, combined with the built-in GPS positioning chip of the mobile phone. When the vehicle enters the 300-meter range around the parking lot, it automatically wakes up and receives the idle parking space information pushed by the server (presented in the form of map annotations, accurate to the parking space number and coordinates, and the coordinates use the internationally accepted WGS84 coordinate system to ensure accurate positioning on the map) and navigation routes (based on A* algorithm optimization, the algorithm comprehensively considers factors such as the distance between nodes and the travel cost during the search process, and selects the optimal path by continuously evaluating the heuristic function values ​​of the surrounding nodes. The navigation route contains detailed turning prompts, distance estimates, and estimated arrival time information, which facilitates car owners to accurately go to the target parking space).

[0027] At the same time, the parking lot entrance display screen is connected to the server through RS-232 or RS-485 serial port, and displays the total number of parking spaces in the parking lot, the number of free parking spaces (displayed in eye-catching numbers and graphical progress bars, for example, the number of free parking spaces is indicated by a green progress bar, and the number of occupied parking spaces is indicated by a red progress bar. The length of the progress bar is dynamically adjusted according to the proportion of the number of parking spaces, so that car owners can clearly understand the overall parking situation of the parking lot at a glance), charging standards (listed in a table in the form of detailed prices by time period and vehicle type, such as 5 yuan per hour for small cars during the day on weekdays, 2 yuan per hour at night, and large cars are increased by a certain proportion on this basis), etc., so that car owners can make parking decisions quickly. The parking space guidance indicator receives server instructions through the ZigBee wireless communication module, and switches colors according to the parking space status. The idle is green (wavelength 520-570nm), and the occupied is red (wavelength 620-760nm). The flashing frequency is dynamically adjusted, and the idle is always on, the occupied is slow flashing (1-2 times per second), and the occupied is fast flashing (3-5 times per second), guiding car owners to find parking spaces in an intuitive and eye-catching way.

[0028] Through the sophisticated processing and multi-factor feedback mechanism of the application layer in step S3, the system accurately and promptly conveys the parking space status determination results and related information to the corresponding users, realizing the intelligent and humanized service of the entire parking space remote monitoring and management system, and significantly improving the management efficiency of the parking lot and the user's parking experience.

[0029] Working principle: The parking space remote monitoring and management system based on RFID technology of the present invention realizes efficient and stable operation through the coordinated work of the perception layer, transmission layer and application layer. Its working principle is as follows: In the perception layer, the ultra-high frequency RFID reader-writer transmits a radio frequency signal of a specific frequency band based on the principle of electromagnetic induction, which stimulates the RFID tag that complies with the EPC Gen2 standard carried by the vehicle to resonate and feedback an electromagnetic signal carrying the vehicle identification information. The reader-writer is connected to the relevant high-speed data acquisition circuit through the SPI interface to capture the signal at a rate of 50Mbps to obtain the vehicle identification information. At the same time, it relies on the built-in precise calibration module to compare the standard signal source in real time to fine-tune the parameters, and control the RSSI measurement error within the range of ±2dBm to assist in judging the distance between the vehicle and the reader-writer, etc., and its built-in spectrum monitoring module can sense the frequency band interference status with a millisecond response speed, and switch the frequency band according to the threshold set by multi-dimensional considerations to ensure the stability and continuity of tag reading in complex electromagnetic environments; the high-performance piezoelectric ceramic transducer of the ultrasonic sensor is excited by the driving circuit to the vehicle The 40kHz ultrasonic pulse is transmitted directionally in the space above the parking space. After being reflected by obstacles, the receiving circuit captures the echo. The ADC conversion module converts the sound wave reflection time into a distance value to infer the vertical occupancy of the parking space. Its built-in intelligent temperature compensation circuit uses thermistors to sense the ambient temperature and dynamically adjusts the sound wave parameters based on the temperature-precision calibration table, so that the fluctuation amplitude of the measured distance under different temperature conditions is controlled within the threshold of ±0.5cm to provide vertical occupancy information support for the judgment of the parking space status. The geomagnetic sensor relies on a high-permeability multi-layer magnetic shielding cover made of Permalloy to reduce the external stray magnetic field by more than 90%, and uses the characteristics of the vehicle's iron-containing parts to disturb the geomagnetic field. The internally integrated 12-bit ADC converts the change in geomagnetic field intensity into a digital signal output, and transmits the three-axis magnetic field intensity data to the STM32F4 microcontroller every 0.1 second through the I2C interface. The microcontroller stores and timestamps the data according to the preset quantization accuracy standard to ensure the synchronization and integrity of the data. Even in the external magnetic field interference scenario, it can accurately detect and output accurate data to provide a reliable basis for the horizontal occupancy judgment of the parking space.

[0030] In the transmission layer, industrial Ethernet uses Category 6 or Category 6a twisted pair cables as the transmission medium in accordance with the IEEE802.3 standard. Each data acquisition node is connected to the Ethernet switch through an RJ45 interface. The switch uses packet switching technology to encapsulate sensor data into Ethernet frames and forwards them according to MAC address routing. It intelligently learns MAC addresses to quickly and accurately forward data to avoid conflicts and congestion. When vehicles frequently enter and exit and sensors collect data and transmit it back at a high frequency, it can ensure that data is quickly transmitted to the management server according to established routing rules to avoid loss or delay, providing accurate data for parking space management decisions; Wi-Fi 6 wireless network uses OFDMA technology to intelligently allocate channels in the 2.4GHz and 5GHz frequency bands according to channel resource occupancy, divides wireless channels into multiple subcarriers and allocates them according to terminal device requirements to achieve parallel communication of multiple devices, and multiple wireless access points. The access point uses Mesh networking to automatically establish a reliable wireless link to form a seamless roaming network environment. When the mobile terminal accesses, it can quickly respond to the allocation of channel resources, and can also dynamically adjust the bandwidth to maintain good network performance to meet diverse access needs and ensure stable and smooth data transmission. The 5G network plays an important role in data traffic peak areas such as parking lot entrances and exits with its low latency and high bandwidth advantages. By integrating with the operator network and using network slicing technology to open up dedicated slices to ensure high priority for data transmission in this system, it can quickly undertake and efficiently transmit data to ensure real-time and complete data when required for large-capacity data transmission. The server-side intelligent routing algorithm monitors the status of each link at extremely short time intervals. According to key parameters such as bandwidth utilization, when a link fails or performance deteriorates, the data transmission is switched to the optimal backup link according to preset decision logic to ensure data transmission continuity.

[0031] In the application layer, in terms of parking space status judgment based on the multivariate data fusion algorithm, the parking lot management server is equipped with a high-performance multi-core processor and a large-capacity memory to run the customized parking space management software. Its core algorithm is the multivariate data deep fusion algorithm of the improved Bayesian network and DS evidence theory. First, in the data receiving link, the data from each sensor in the transmission layer is parsed and verified according to the strict format and protocol, and stored in the cache area to clean up the abnormal data. Then, a Bayesian network with 4 layers of nodes is constructed. Each node establishes a directed connection based on physical principles and practical experience. A large amount of historical parking lot operation data is used to use the maximum likelihood estimation method to learn the conditional probability distribution between nodes and construct a conditional probability table. After the Bayesian network preliminarily infers the parking space status judgment results of each sensor, these results are converted into independent evidence sources. The basic probability assignment is reasonably allocated according to multi-dimensional factors such as the sensor's historical accuracy, current working status, and data consistency. Then, the DS synthesis rule is used to fuse the BPA of each sensor multiple times to overcome the limitations of single sensor judgment and improve the accuracy of the vehicle. Accuracy of parking space status judgment; In terms of information feedback and display mechanism, for parking lot managers, the server-side database stores various types of information in different data tables. Managers can obtain statistical reports covering key indicators such as parking space utilization rate, income, and failure rate through query statements to understand the operation status and make decision adjustments. For the majority of car owners, mobile APP develops integrated functional modules based on relevant platforms, and uses push notifications and GPS positioning chip functions. When a vehicle enters the parking lot within 300 meters, it automatically receives information on available parking spaces and navigation routes optimized based on the A* algorithm. The parking lot entrance display screen connects to the server through a serial port and receives data according to the communication protocol to display the total number of parking spaces, the number of available parking spaces, charging standards and other information in real time. The parking space guide indicator light receives server instructions through the ZigBee wireless communication module to switch colors and adjust the flashing frequency according to the parking space status, guiding car owners to find parking spaces in an intuitive way, ultimately realizing the intelligent and humanized services of the entire system, improving parking lot management efficiency and user parking experience.

[0032] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A parking space remote monitoring and management system based on RFID technology, characterized in that: include: The sensing layer consists of UHF RFID readers, ultrasonic sensors, and geomagnetic sensors distributed in each parking space in the parking lot; The UHF RFID reader is used to identify the RFID tag carried by the vehicle to obtain the vehicle identification information; The ultrasonic sensor detects the distance of obstacles above the parking space through the principle of sound wave reflection to determine the vertical occupancy of the parking space; The geomagnetic sensor is used to detect the change of magnetic field intensity in the parking area according to the disturbance of the vehicle's iron-containing parts to the geomagnetic field to determine the horizontal occupancy of the parking space; The transport layer uses a combination of wired and wireless methods, including industrial Ethernet, Wi-Fi, and 4G / 5G networks, to transmit the data collected by the perception layer to the application layer; The application layer includes the parking lot management server, database, mobile phone APP for car owners, and display terminals in the parking lot; The parking lot management server is used to receive data from the transmission layer and run data fusion, analysis and decision-making algorithms to perform intelligent management of parking spaces; The database is used to store vehicle information, parking space status historical data, and user information, and the mobile phone APP is used to provide parking space query, reservation, and navigation services for car owners; The display terminal is used to display the overall situation of parking spaces in the parking lot in real time to guide car owners to park.

2. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The UHF RFID system has a built-in large-capacity storage chip that can record the vehicle's license plate number, model, owner's contact information and entry time information, and the chip has an adaptive tuning function.

3. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The temperature compensation circuit built into the ultrasonic sensor automatically adjusts the measurement parameters according to the ambient temperature.

4. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The geomagnetic sensor is a three-axis geomagnetic sensor with a measurement sensitivity of 0.1nT. A high-precision ADC is integrated inside the sensor to convert the magnetic field strength into a digital signal output, and the sensor is equipped with a multi-layer magnetic shielding cover outside.

5. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: In the sensing layer, the RFID reader is installed on one side of the parking space at a height of 2-2.5 meters from the ground, and the antenna is tilted 30°-45° toward the parking space; the ultrasonic sensor is installed 0.8-1.2 meters directly above the parking space; the geomagnetic sensor is buried in the center of the parking space ground at a depth of 5-10 centimeters, and is covered with a layer of electromagnetic shielding and heat insulation composite layer.

6. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The parking lot management server adopts a multi-data deep fusion algorithm based on improved Bayesian network and DS evidence theory, including a data acquisition and preprocessing module, a Bayesian network reasoning module, and a DS evidence theory fusion module; The data acquisition and preprocessing module uses an STM32F4 microcontroller to build a circuit, synchronously triggers each sensor to collect data through a timer, and preprocesses the collected data using a median filter, a Kalman filter, and a CRC check algorithm; The Bayesian network reasoning module constructs a Bayesian network containing four layers of nodes, the first layer is the sensor observation node, the second layer is the environmental impact node, the third layer is the parking space status intermediate node, and the top layer is the parking space real status node; The DS evidence theory fusion module performs DS evidence theory fusion after the Bayesian network inference obtains the preliminary judgment of each sensor on the parking space status.

7. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The transmission layer adopts multi-link redundant communication, deploys industrial Ethernet as the backbone network in the parking lot, connects various data collection nodes with the management server, and deploys Wi-Fi6 wireless network to cover the entire parking lot.

8. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: In the application layer, the parking lot management server uses RabbitMQ message queue technology to push information including parking space status updates and vehicle entry and exit records to relevant terminals.

9. The parking space remote monitoring and management system based on RFID technology according to claim 1 is characterized in that: The system also includes an intelligent self-repair and adaptive adjustment module, which starts a machine learning-based intelligent diagnosis program every 3 minutes, collects data from each sensor in the past 10 minutes, including features such as measurement values, data fluctuation range, and signal strength, and inputs it into a pre-trained decision tree model to determine whether the sensor is working properly.

10. The parking space remote monitoring and management system based on RFID technology according to claim 1, characterized in that: The system also includes a remote collaborative operation and maintenance module, which uses a microservice architecture to perform cluster analysis and correlation analysis on the data of multiple parking lots of the same type.

Images